The vibrational spectrum of Mg2SiO4 olivine was calculated at the Γ point by using the periodic ab initio CRYSTAL program. An all electron localized Gaussian-type basis set and the B3LYP Hamiltonian were employed. The full set of frequencies (35 IR active, 36 Raman active, 10 "silent" modes) was computed and compared to experimental data from different sources (four for IR and four for Raman). A generally good agreement is observed with experiment (the mean absolute difference ranging from 7 to 10 cm-1 for the various sets), when some of the experimental frequencies, whose attribution is uncertain or appears to be affected by large errors, are not taken into account. A small number of observed peaks are not consistent with calculated frequencies, and a few theoretical peaks do not correspond to measured values. The implications are discussed in detail. The full set of modes are characterized using different tools, namely isotopic substitution, direct inspection of the eigenvectors and graphical representation, so as to obtain a consistent mode assignment. © Springer-Verlag 2006.
Noel, Y., Catti, M., D'Arco, P., Dovesi, R. (2006). The vibrational frequencies of forsterite Mg2SiO4; an all-electron ab initio study with the CRYSTAL code. PHYSICS AND CHEMISTRY OF MINERALS, 33(6), 383-393 [10.1007/s00269-006-0085-y].
The vibrational frequencies of forsterite Mg2SiO4; an all-electron ab initio study with the CRYSTAL code
CATTI, MICHELE;
2006
Abstract
The vibrational spectrum of Mg2SiO4 olivine was calculated at the Γ point by using the periodic ab initio CRYSTAL program. An all electron localized Gaussian-type basis set and the B3LYP Hamiltonian were employed. The full set of frequencies (35 IR active, 36 Raman active, 10 "silent" modes) was computed and compared to experimental data from different sources (four for IR and four for Raman). A generally good agreement is observed with experiment (the mean absolute difference ranging from 7 to 10 cm-1 for the various sets), when some of the experimental frequencies, whose attribution is uncertain or appears to be affected by large errors, are not taken into account. A small number of observed peaks are not consistent with calculated frequencies, and a few theoretical peaks do not correspond to measured values. The implications are discussed in detail. The full set of modes are characterized using different tools, namely isotopic substitution, direct inspection of the eigenvectors and graphical representation, so as to obtain a consistent mode assignment. © Springer-Verlag 2006.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.